Journal Articles
  • Dieversity Priors for Learning Early Visual Features   
         Hanchen Xiong, Antonio Rodr¨ªguez-S¨¢nchez, Sandor Szedmak,  Justus Piater
         [DOI], Frontiers in Computational Neuroscience, 2015 [Data & Code]
  • Learning Undirected Graphical Models using Persistent Sequential Monte Carlo   
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         (Under review) Machine Learning Journal
  • Structural bootstrapping-A novel, generative mechanism for faster and more efficient acquisition of action knowledge.   
         Florentin W&oumlrg&oumltter, Christopher Geib, Minija Tamosiunaite, Eren Erdal Aksoy, Justus Piater, Hanchen Xiong, Alex Ude, Bojan Nemec, Dirk Kraft,
         Nobert Kr&uumlger, Mirko W&aumlchter, Tamim Asfour
         [DOI], IEEE Transactions on Autonomous Mental Development, 2015 (Volume 7, Issue 2).
  • Scalable, Accurate Image Annotation with Joint SVM and Output Kernels  
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         [DOI], Neurocomputing Journal, 2015,  [PDF] [bib]
  • A New Binary-Tree-Based Algorithm for XML Data Model Extraction
         Qing Lei,  Hanchen Xiong
         Computer Engineering and Design 2009 (13) [bib]

Conference Proceeding Papers

  • Learning Missing  Edges via Kernels in Partially-known Graphs
        Senak Krivic, Sandor Szedmak, Hanchen Xiong, Justus Piater
        23rd European Symposium on Artificial Neural Network (ESANN15), to appear

  • Towards Maximum Likelihood: Learning Undirected Graphical Models using Persistent Sequential Monte Carlo
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         6th Asian Confernce on Machine Learning (ACML14) (oral presentation),  Vietnam, 2014
         JMLR W&CP(DOI) [PDF] [Slides] [Poster] [bib] , Best Paper Award.
         Invited to "Machine Learning" Journal with an extended version
  • Towards Sparsity and Selectivity: Bayesian Learning of Restricted Boltzmann Machine for Early Visual Features
         Hanchen Xiong, Sandor Szedmak,  Antonio Rodr¨ªguez-S¨¢nchez, Justus Piater
         24th International Conference on Artificial Neural Networks (ICANN14), Hamburg, Germany, September, 2014
         Springer-Verlag(DOI) [PDF] [Poster] [bib]
  • Joint SVM for Accurate and Fast Image Tagging
        Hanchen Xiong, Sandor Szedmak, Justus Piater
        22nd European Symposium on Artificial Neural Network (ESANN14)  (oral presentation), Bruges, Belgium, April, 2014
        ESANN(DOI) [PDF] [Slides] [bib]
        Invited to "Neurocomputing" Journal with an extended version
  • 3D Object Class Geometry Modeling with Spatial Latent Dirichlet Markov Random Fields
        Hanchen Xiong, Sandor Szedmak, Justus Piater
        35th German Conference on Pattern Recognition (GCPR13), Saarbruken, Germany, September, 2013
        Sringer-Verlag(DOI) [PDF] [Poster] [Data & Code] [bib]
  • Homogeneity Analysis for Object-Action Relation Reasoning in Kitchen Scenarios
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         2nd Workshop on Machine Learning for Interactive Systems (MLIS13) (oral presentation), Beijing, China, August, 2013
         ACM(DOI) [PDF] [Slides]  [Data & Code] [bib]
  • A Study of Point Cloud Registration with Probability Product Kernel Functions
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         2013 International Conference on 3D Vision (3DV13),  Seattle, Washington, USA,  June, 2013
         IEEE(DOI) [PDF] [Supplementary Material] [bib]
  • Efficient, General Point Cloud Registration with Kernel Feature Maps
         Hanchen Xiong, Sandor Szedmak, Justus Piater
         10th International Conference on Computer and Robotic Vision (CRV13)  (oral presentation), Regina, Canada, May,  2013
         IEEE(DOI) [PDF] [Slides] [Data & Code] [bib]

Peer-Reviewed Short Papers

  • Implicit Learning of Simpler Output Kernels for Multi-label prediction
          Hanchen Xiong, Sandor Szedmak, Justus Piater
          NIPS workshop on Representation and Learning Methods for Complex Outputs (NIPS_RLCO14) , Montreal , Canada,  2014
          [PDF] [Slides] [Poster] [bib]
  • Comparing Binary Hamiltonian Monte Carlo and Gibbs Sampling for Training Discrete MRFs with Stochastic Approximation
          Hanchen Xiong, Sandor Szedmak, Justus Piater
         17th International COnference on Artificial Intelligence and Statistics (AISTATS14) , Reykjavik, Iceland, April,  2014
          [PDF] [Poster] [bib]

Technical Reports
  • Learning Shapes as Directed Closed Surfaces
        Sandor Szedmak,  Hanchen Xiong, Justus Piater,  December, 2011
        Institute of Computer Science, University of Innsbruck, [PDF] [bib]

These
  • Inference, Learning and Optimization on Structured Domains: Methods and Applications
         Hanchen Xiong, Supervisor: Justus Piater
         Ph.D Dissertation, Institute of Informatics, University of Innsbruck, 2015. [PDF] [bib]
  • Investigate The Use of Random Finite Set to Track Quantum Dots  (Distinction)
         Hanchen Xiong, Supervisor: Simion Julier
         Master Thesis, Department of Computer Science, University College London, 2010.